CODEWITHSUNDEEP
machine-learningneural-networkprotocol-buffersdeep-learningcaffe
BHAV247
BHAV247

Reputation: 69

caffe | Error parsing ... String literals cannot cross line boundaries

I would like to fine tune bvlc_reference_caffenet model on my dataset.
I did the following steps:

  1. I created lmdb files for the new dataset I had.

  2. I downloaded the Pretrained bvlc_reference_caffenet.caffemodel.

  3. I changed the last layer i.e "fc8", fully connected layer:
    Took the train_val.prototxt and changed the last layer "fc8" to "fc8_tune".

  4. I changed num_output to the right number of output classes trying to predict (i.e 2).

  5. I changed solver.prototxt according my data

  6. I run 

    $TOOLS/caffe train --solver=solver.prototxt --weights=bvlc_reference_caffenet.caffemodel

but I got the following error

[libprotobuf ERROR google/protobuf/text_format.cc:245] Error parsing text-format caffe.NetParameter: 363:19: String literals cannot cross line boundaries.
[libprotobuf ERROR google/protobuf/text_format.cc:245] Error parsing text-format caffe.NetParameter: 363:18: Message type "caffe.LayerParameter" has no field named "fc8_tune".
F0531 17:13:33.284981  3670 upgrade_proto.cpp:79] Check failed: ReadProtoFromTextFile(param_file, param) Failed to parse NetParameter file: /caffe-master/models/bvlc_reference_caffenet/train_val.prototxt

can anyone help me in solving this error?

Upvotes: 1

Views: 1933

Answers (1)

Shai
Shai

Reputation: 114786

You forgot to close a quotation mark (") in your prototxt file.
According to the error message

Error parsing text-format caffe.NetParameter: 363:19:

you should look around line number 363.

Upvotes: 2

Related Questions